Will computers ever become thoughtful enough to earn personhood? Fifty years of experimentation and debate have brought researchers no closer to a consensus on the issue, perhaps because it crams a vexing scientific question-What is a thought? -up against an even stickier philosophical and ethical one: What is a person? In The Cambridge Quintet, physicist-writer John L. Casti enlists five of the century’s most distinguished intellectuals as the dramatis personae in a heated, hypothetical dinner conversation exploring this fault line.
Under Casti’s direction, the characters add as much to the fun as the subject matter. The dinner is set at England’s Cambridge University in 1949, and the host, appropriately, is the physicist and novelist C.P. Snow. Arguing for the plausibility of artificial intelligence is Alan Turing, the introverted mathematician and World War II codebreaker. Turing outlines for his colleagues the proposition (now famously known as the Turing test) that any machine with conversational skills indistinguishable from a human’s must be regarded as intelligent, no matter how “programmed” this behavior may be.
Across the table, and across a vast ideological gulf, from Turing is the philosopher Ludwig Wittgenstein. At dinner, as in his writings, Wittgenstein posits that the essential medium of thought is language, and that language is based on culturally shared rules. With no culture or experiences of their own, Wittgenstein declares firmly, even machines capable of passing Turing’s test would not be persons, and would have no genuine understanding of the words they used; “They may have machine dreams,” he fumes, “but those dreams are as far from being the dreams of a human as a steam shovel is from being the college gardener digging in the courtyard.”
Don’t settle for half the story.
Get paywall-free access to technology news for the here and now.